Re: [patch] revert: [NET]: Fix races in net_rx_action vs netpoll

[Ingo Molnar <mingo@elte.hu>]

* Olaf Kirch <olaf.kirch@oracle.com> wrote:

> Can you try what happens if you change netif_rx_complete to something 
> like this:
> 
> 	if (test_bit(__LINK_STATE_POLL_LIST_FROZEN, &dev->state)) {
> 		dev->quota = dev->weight;
> 		return;
> 	}
> 
> This is just a hack to make sure that we don't go to insanely negative 
> quotas while sending packets through netpoll.

i've done the patch below, but it did not change the timeouts nor did it 
solve the 'no network' problem. netconsole output hung earlier as well.

i can try other things too. (it would be best if you sent me 
test/debug-patches, instead of letting me hack in the changes - that's 
the surest way of ensuring that i test exactly what you intended.)

	Ingo

------------->
Index: linux/include/linux/netdevice.h
===================================================================
--- linux.orig/include/linux/netdevice.h
+++ linux/include/linux/netdevice.h
@@ -1007,6 +1007,10 @@ static inline int netif_rx_reschedule(st
  */
 static inline void __netif_rx_complete(struct net_device *dev)
 {
+	if (test_bit(__LINK_STATE_POLL_LIST_FROZEN, &dev->state)) {
+		dev->quota = dev->weight;
+		return;
+	}
 	BUG_ON(!test_bit(__LINK_STATE_RX_SCHED, &dev->state));
 	list_del(&dev->poll_list);
 	smp_mb__before_clear_bit();